As pointed out in the Lacy, et al. article (Winter, 2001) in order for accurate conclusions to be drawn about any testable situation there must exist an adequate sample size as well as employing an appropriate sampling structure. Further, according to the authors a sample drawn at random is an unbiased sample in the sense that no member of the population has a greater chance of being selected than any other member. At the same time non-random samples are, unfortunately, over or under-representations of a population and have a tendency to skew measurement results. There, are, however, research situations wherein randomization is not possible but this does not necessarily negate the research effort. The implication behind non-randomization is that the results of the study, whether they are differences, relationships, or effects, cannot be levied against a greater...

The results must be solely directed toward the sample itself. In other words, generalizations that a population will mirror the sample results is not possible - even though many researchers commit this particular research protocol transgression.
Not only is the type of sampling important to effective and credible research but the size of the sample selected is also important; and this particular research tenet or rule is the most difficult to achieve. The ruff and ready rule with respect to sample size is to garner the largest size possible. Whenever a mean, a percentage, or any other statistic is computed from a sample, a population value is being estimated (Wilkerson & Olson, 1997). On this population estimation the established testable null hypothesis is either accepted or rejected. The primary question the research